CN107563330A - A kind of horizontal tilt car plate antidote in monitor video - Google Patents
A kind of horizontal tilt car plate antidote in monitor video Download PDFInfo
- Publication number
- CN107563330A CN107563330A CN201710784790.0A CN201710784790A CN107563330A CN 107563330 A CN107563330 A CN 107563330A CN 201710784790 A CN201710784790 A CN 201710784790A CN 107563330 A CN107563330 A CN 107563330A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- straight line
- top edge
- mtd
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
- Character Input (AREA)
Abstract
The invention provides the horizontal tilt car plate antidote in a kind of monitor video, belong to automatic Recognition of License Plate field, it is therefore intended that the horizontal tilt car plate navigated in monitor video is corrected, pre-processed for Car license recognition.License plate image is obtained from monitor video, extracts each character zone therein；It is fitted to obtain three straight lines, including top edge straight line, central line, lower edge straight line respectively using the top edge midpoint of multiple character zones, central point, lower edge midpoint, and calculates the horizontal tilt angle corresponding to every straight line；Finally, three respective weights of straight line are calculated according to the error of fitting of straight line, merges the inclination angle of three straight lines, calculate the optimum level angle of inclination of car plate, and carry out rectification.Compared with prior art, it is higher according to the optimum level angle of inclination accuracy rate of the invention detected, and calculating speed is fast, has obvious advantage to the incomplete car plate of frame, field is corrected suitable for car plate.
Description
Technical field
The present invention relates to the Car license recognition in image procossing, video monitoring and safety-security area, especially intelligent transportation field
Method.
Background technology
Nowadays, with the continuous improvement of people's living standards, motor vehicle growing day by day to traffic department supervision with
Management brings stern challenge, and intelligent transportation system, which seems, to be even more important, and identity of the car plate as vehicle so that from
Dynamic Car license recognition (ALPR) is the key technology in intelligent transportation system, plays and focuses in fields such as vehicle management, monitoring violating the regulations
Act on.
However, due to the position relationship of camera in actual monitored scene, in the video frame images of acquisition there is angle in car plate
Tilt problem is spent, the Car license recognition in later stage can be had an impact.
At present, car plate angle correction mainly includes following several method：(1) method based on straight-line detection, this method are led to
The angle of inclination that detection car plate frame straight line obtains car plate is crossed, so as to be corrected to car plate, but this method is to having interference, obscuring
Car plate it is ineffective；(2) method of feature based analysis, such as principal component analysis, but this method is vulnerable to non-character region
Interference；(3) method based on the method for exhaustion, this method is by constantly by the rotation of car plate progress certain angle, then obtaining corresponding
Characteristic value, compare to obtain the angle corresponding to best features value, as angle of inclination, this method robustness is preferable, but calculate
Measure larger.
And for the vehicle in monitoring scene, there is phenomena such as incomplete frame, ambient interferences in the car plate navigated to, to sum up institute
State, there is an urgent need to propose the horizontal tilt car plate antidote in a kind of monitor video at present.
The content of the invention
The present invention proposes the horizontal tilt car plate antidote in a kind of monitor video, for realizing car plate horizontal tilt angle
Degree fast and accurately detects and VLP correction.
The technical solution adopted for the present invention to solve the technical problems is：
A kind of horizontal tilt car plate antidote in monitor video, including：
Color License Plate image in S1, acquisition monitoring scene, gray processing, binaryzation are carried out to Color License Plate image, opens fortune
Calculation and connected domain contour detecting algorithm, obtain the rectangular bounding box of multiple connected domain profiles, according to the geometrical property of characters on license plate
Screened, remove the rectangular bounding box of interference, satisfactory rectangular bounding box is defined as character zone；
S2, to the car plate containing multiple character zones, least square method is utilized according to the top edge midpoint of each character zone
Fitting obtains a top edge midpoint straight line, and one is obtained using least square fitting according to the central point of each character zone
Central point straight line, and obtained according to the lower edge midpoint of each character zone using least square fitting in a lower edge
Point straight line, the horizontal tilt angle of three straight lines is calculated respectively；
If the distance for the top edge midpoint straight line that S3, top edge midpoint obtain to fitting is bigger, top edge midpoint straight line
Error of fitting it is bigger, the confidence level of top edge midpoint straight line is also lower.Therefore can be according to the top of multiple character zones
Edge midpoint determines the confidence level of top edge midpoint straight line to the average distance of top edge midpoint straight line, and then calculates top edge midpoint
The weight of straight line.The weight of central point straight line and lower edge midpoint straight line can similarly be calculated；
S4, inclined according to top edge midpoint straight line, the weight of central point straight line and lower edge midpoint straight line and respective level
Rake angle is merged, and calculates the optimum level angle of inclination of car plate；
S5, the car plate optimum level angle of inclination according to determination, horizontal tilt correction is carried out to car plate.
Further, the step S1, including：
S11, input color image size be Height × Width, gray processing is carried out to it, obtain gray-scale map f (x,
Y), in order to remove the interference at edge, gray-scale map middle size is taken to calculate threshold value T for height × width regions, calculation formula is such as
(1) binaryzation, is carried out to gray-scale map with this threshold value T and obtains bianry image, carries out opening operation, connected domain wheel successively to bianry image
Exterior feature detection, finally give the rectangular bounding box of multiple connected domain profiles；
Car plate is fuzzy in the geometrical property of character and actual scene in S12, combined standard license plate image, incomplete, car plate word
It is R to accord with factor, the aspect ratio ranges for setting characters on license plate such as adhesionw/h, the area of characters on license plate is Rarea, according to characters on license plate
Characteristic rejects undesirable rectangular bounding box, and specific rules are the ratio of width to height of rectangular bounding box in Rw/hIn the range of, rectangular edges
Boundary's frame occupied area ratio is less than Rarea, all rectangular bounding box to be analyzed, rejecting does not meet this regular rectangular bounding box,
Satisfactory M rectangular bounding box is defined as character zone.
Further, the step S2, including：
S21, the position [x according to i-th (i=1 .., M) individual character zonei,yi,wi,hi] calculate point coordinates in top edgeCenter point coordinatePoint coordinates in lower edgeWherein (xi,yi) represent i-th of character
The top left co-ordinate in region, (wi,hi) represent i-th of character zone wide and high, calculation formula such as (2)
S22, to M character zone repeat step S21, obtain point sequence in top edgeCenter point sequencePoint sequence in lower edgeLeast square fitting is carried out to point sequence in top edge, obtained in top edge
Point straight line Lup:yup=kupx+bup, least square fitting is carried out to center point sequence, obtains central point straight line Lmid:ymid=kmidx
+bmid, least square fitting is carried out to point sequence in lower edge, obtains lower edge midpoint straight line Ldown:ydown=kdownx+bdown,
And the horizontal tilt angle α of this three straight lines is calculated according to straight slopeup、αmid、αdown, calculation formula such as (3)
αl=arctankl l∈{up,mid,down} (3)
Further, the step S3, including：
S31, for top edge midpoint straight line, ask corresponding to average departure of the point sequence to top edge midpoint straight line in top edge
FromFor central point straight line, corresponding center point sequence is sought to the average distance of central point straight lineFor in lower edge
Point straight line, ask corresponding to average distance of the point sequence to top edge midpoint straight line in lower edgeCalculation formula such as (4)
WhereinRepresent sequence PointlIn i-th point arrive straight line Ll:yl=klx+blDistance, calculation formula such as (5)
S32, the confidence level for defining top edge midpoint straight line, central point straight line and lower edge midpoint straight line are respectively λup、
λmid、λdown, calculation formula such as (6)
And the weight for defining top edge midpoint straight line, central point straight line and lower edge midpoint straight line is respectively wup、wmid、
wdown, calculation formula such as (7)
Further, the step S4, it is specially：
The inclined angle alpha of top edge midpoint straight line, central point straight line and lower edge midpoint straight line is calculated in above-mentioned stepsup、
αmid、αdown, and respective weight wup、wmid、wdown, the optimum level angle of inclination for thus calculating car plate is
α=wupαup+wmidαmid+wdownαdown (8)
Further, the step S5, it is specially：
With license plate image central pointFor pivot, the above-mentioned optimum level inclination angle being calculated
Degree α is the anglec of rotation, rotates license plate image, you can obtain final horizontal tilt correction result.
Beneficial effect
Horizontal tilt car plate antidote in a kind of monitor video of the present invention, compared with the existing methods, advantage exists
In：
(1) can solve the angle problem of inclination license plate image slightly hazy, that edge is incomplete；
(2) straight-line fitting techniques are utilized, computation complexity is relatively low, can quickly obtain optimum level angle of inclination；
(3) a plurality of straight line is utilized, and is merged with reference to straight line error, robustness is preferable.
The car plate horizontal tilt angle that horizontal tilt car plate antidote in a kind of monitor video of the present invention obtains
Degree, fast and accurately rectification can be carried out to car plate, suitable for Car license recognition field.
Brief description of the drawings
Fig. 1：The schematic flow sheet of horizontal tilt car plate antidote in a kind of monitor video of the present invention；
Specific embodiment
Below in conjunction with the accompanying drawings and technical solution of the present invention is described in detail preferred embodiment.
Schematic flow sheet such as Fig. 1 institutes of horizontal tilt car plate antidote in a kind of monitor video provided by the invention
Show, including：
Color License Plate image in S1, acquisition monitoring scene, gray processing, binaryzation are carried out to Color License Plate image, opens fortune
Calculation and connected domain contour detecting algorithm, obtain the rectangular bounding box of multiple connected domain profiles, according to the geometrical property of characters on license plate
Screened, remove the rectangular bounding box of interference, satisfactory rectangular bounding box is defined as character zone；
S2, to the car plate containing multiple character zones, least square method is utilized according to the top edge midpoint of each character zone
Fitting obtains a top edge midpoint straight line, and one is obtained using least square fitting according to the central point of each character zone
Central point straight line, and obtained according to the lower edge midpoint of each character zone using least square fitting in a lower edge
Point straight line, the horizontal tilt angle of three straight lines is calculated respectively；
S3, the distance of top edge midpoint straight line obtained to fitting due to top edge midpoint are bigger, top edge midpoint straight line
Error of fitting it is bigger, the confidence level of top edge midpoint straight line is also lower.Therefore can be according to the top of multiple character zones
Edge midpoint determines the confidence level of top edge midpoint straight line to the average distance of top edge midpoint straight line, and then calculates top edge midpoint
The weight of straight line.The weight of central point straight line and lower edge midpoint straight line can similarly be calculated；
S4, inclined according to top edge midpoint straight line, the weight of central point straight line and lower edge midpoint straight line and respective level
Rake angle is merged, and calculates the optimum level angle of inclination of car plate；
S5, the car plate optimum level angle of inclination according to determination, horizontal tilt correction is carried out to car plate.
Further, the step S1, including：
S11, input color image size be Height × Width, gray processing is carried out to it, obtain gray-scale map f (x,
Y), in order to remove the interference at edge, the gray-scale map middle size is taken to be
Region calculates gray threshold T, and calculation formula such as (1), with this threshold value T carries out binaryzation to gray-scale map, bianry image obtained, to two
Value image carries out opening operation, connected domain contour detecting successively, finally gives the rectangular bounding box of multiple connected domain profiles；
The ratio of width to height of character is 2 in S12, standard license plate image, it is contemplated that car plate in actual scene is fuzzy, it is incomplete because
Element, the ratio of width to height R of characters on license plate is setw/h=1~3, it is contemplated that the factor such as Characters Stuck in license plate image, characters on license plate is set
The area R in region shared region in license plate imagearea=Height × Width/2, all rectangular bounding box are analyzed,
Specific rules are the ratio of width to height of rectangular bounding box in Rw/hIn the range of, the area of rectangular bounding box is less than Rarea, picked according to this rule
Except the rectangular bounding box for not meeting character property, 6 rectangular bounding box for meeting characteristic are defined as character zone.
Further, the step S2, including：
S21, the position [x according to i-th (i=1 .., 6) individual character zonei,yi,wi,hi] calculate point coordinates in top edgeCenter point coordinatePoint coordinates in lower edgeWherein (xi,yi) represent i-th of character
The top left co-ordinate in region, (wi,hi) represent i-th of character zone wide and high, calculation formula such as (2)
S22, to 6 character zone repeat step S21, obtain point sequence in top edgeCenter point sequencePoint sequence in lower edgeLeast square fitting is carried out to point sequence in top edge, obtains top edge midpoint
Straight line Lup:yup=kupx+bup, least square fitting is carried out to center point sequence, obtains central point straight line Lmid:ymid=kmidx+
bmid, least square fitting is carried out to point sequence in lower edge, obtains lower edge midpoint straight line Ldown:ydown=kdownx+bdown,
And the horizontal tilt angle α of this three straight lines is calculated according to straight slopeup、αmid、αdown, calculation formula such as (3)
αl=arctankl l∈{up,mid,down} (3)
Further, the step S3, including：
S31, for top edge midpoint straight line, ask corresponding to average departure of the point sequence to top edge midpoint straight line in top edge
FromFor central point straight line, corresponding center point sequence is sought to the average distance of central point straight lineFor in lower edge
Point straight line, ask corresponding to average distance of the point sequence to top edge midpoint straight line in lower edgeCalculation formula such as (4)
WhereinRepresent sequence PointlIn i-th point arrive straight line Ll:yl=klx+blDistance, calculation formula is such as (5)
S32, define top edge midpoint straight line Lup, central point straight line LmidWith lower edge midpoint straight line LdownConfidence level point
Wei not λup、λmid、λdown, calculation formula such as (6)
The weight of top edge midpoint straight line, central point straight line and lower edge midpoint straight line is respectively wup、wmid、wdown, calculate
Formula such as (7)
Further, the step S4, it is specially：
The inclined angle alpha of top edge midpoint straight line, central point straight line and lower edge midpoint straight line is calculated in above-mentioned stepsup、
αmid、αdown, and respective weight wup、wmid、wdown, the optimum level angle of inclination for thus calculating car plate is
α=αupwup+αmidwmid+αdownwdown (8)
Further, the step S5, it is specially：
With license plate image central pointFor pivot, the above-mentioned optimum level inclination angle being calculated
Degree α is the anglec of rotation, rotates license plate image, you can obtain final horizontal tilt correction result.
Claims (6)
1. the horizontal tilt car plate antidote in a kind of monitor video, it is characterised in that comprise the following steps：
S1, obtain monitoring scene in Color License Plate image, to Color License Plate image carry out gray processing, binaryzation, opening operation and
Connected domain contour detecting algorithm, the rectangular bounding box of multiple connected domain profiles is obtained, carried out according to the geometrical property of characters on license plate
Screening, removes the rectangular bounding box of interference, satisfactory rectangular bounding box is defined as into character zone；
S2, to the car plate containing multiple character zones, least square fitting is utilized according to the top edge midpoint of each character zone
A top edge midpoint straight line is obtained, a center is obtained using least square fitting according to the central point of each character zone
Point straight line, and according to the lower edge midpoint of each character zone using least square fitting to obtain a lower edge midpoint straight
Line, the horizontal tilt angle of three straight lines is calculated respectively；
If the distance for the top edge midpoint straight line that S3, top edge midpoint obtain to fitting is bigger, the plan of top edge midpoint straight line
Conjunction error is bigger, and the confidence level of top edge midpoint straight line is also lower；Therefore can be in the top edge according to multiple character zones
The average distance of point to top edge midpoint straight line determines the confidence level of top edge midpoint straight line, and then calculates top edge midpoint straight line
Weight；The weight of central point straight line and lower edge midpoint straight line can similarly be calculated；
S4, according to top edge midpoint straight line, the weight of central point straight line and lower edge midpoint straight line and respective horizontal tilt angle
Degree is merged, and calculates the optimum level angle of inclination of car plate；
S5, the car plate optimum level angle of inclination according to determination, horizontal tilt correction is carried out to car plate.
2. the method as described in claim 1, it is characterised in that the step S1 includes：
S11, the color image size of input are Height × Width, carry out gray processing to it, obtain gray-scale map f (x, y), be
Remove the interference at edge, take gray-scale map middle size to calculate threshold value T for height × width regions, calculation formula such as (1),
Binaryzation is carried out to gray-scale map with this threshold value T and obtains bianry image, carries out opening operation, the inspection of connected domain profile successively to bianry image
Survey, finally give the rectangular bounding box of multiple connected domain profiles；
<mrow>
<mi>T</mi>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>y</mi>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>H</mi>
<mi>e</mi>
<mi>i</mi>
<mi>g</mi>
<mi>h</mi>
<mi>t</mi>
<mo>-</mo>
<mi>h</mi>
<mi>e</mi>
<mi>i</mi>
<mi>g</mi>
<mi>h</mi>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mo>(</mo>
<mi>H</mi>
<mi>e</mi>
<mi>i</mi>
<mi>g</mi>
<mi>h</mi>
<mi>t</mi>
<mo>+</mo>
<mi>h</mi>
<mi>e</mi>
<mi>i</mi>
<mi>g</mi>
<mi>h</mi>
<mi>t</mi>
<mo>)</mo>
<mo>/</mo>
<mn>2</mn>
</mrow>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>x</mi>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>W</mi>
<mi>i</mi>
<mi>d</mi>
<mi>t</mi>
<mi>h</mi>
<mo>-</mo>
<mi>w</mi>
<mi>i</mi>
<mi>d</mi>
<mi>t</mi>
<mi>h</mi>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mn>2</mn>
</mrow>
<mrow>
<mo>(</mo>
<mi>W</mi>
<mi>i</mi>
<mi>d</mi>
<mi>t</mi>
<mi>h</mi>
<mo>+</mo>
<mi>w</mi>
<mi>i</mi>
<mi>d</mi>
<mi>t</mi>
<mi>h</mi>
<mo>)</mo>
<mo>/</mo>
<mn>2</mn>
</mrow>
</munderover>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>h</mi>
<mi>e</mi>
<mi>i</mi>
<mi>g</mi>
<mi>h</mi>
<mi>t</mi>
<mo>&times;</mo>
<mi>w</mi>
<mi>i</mi>
<mi>d</mi>
<mi>t</mi>
<mi>h</mi>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
Car plate obscures, incomplete, characters on license plate glues in the geometrical property of character and actual scene in S12, combined standard license plate image
Even factor, the aspect ratio range for setting characters on license plate is Rw/h, the area of characters on license plate is Rarea, picked according to characters on license plate characteristic
Except undesirable rectangular bounding box, specific rules are the ratio of width to height of rectangular bounding box in Rw/hIn the range of, rectangular bounding box institute
Account for area ratio and be less than Rarea, all rectangular bounding box are analyzed, rejecting does not meet this regular rectangular bounding box, will meet
It is required that M rectangular bounding box be defined as character zone.
3. the method as described in claim 1, it is characterised in that step S2 includes：
S21, the position [x according to i-th (i=1 .., M) individual character zonei,yi,wi,hi] calculate point coordinates in top edgeCenter point coordinatePoint coordinates in lower edgeWherein (xi,yi) represent i-th of character
The top left co-ordinate in region, (wi,hi) represent i-th of character zone wide and high, calculation formula such as (2)：
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<msubsup>
<mi>x</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>x</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<mfrac>
<msub>
<mi>w</mi>
<mi>i</mi>
</msub>
<mn>2</mn>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<msubsup>
<mi>y</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<msubsup>
<mi>y</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<mfrac>
<msub>
<mi>h</mi>
<mi>i</mi>
</msub>
<mn>2</mn>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<msubsup>
<mi>y</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<msub>
<mi>h</mi>
<mi>i</mi>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
S22, to M character zone repeat step S21, obtain point sequence in top edge
Center point sequencePoint sequence in lower edge
Least square fitting is carried out to point sequence in top edge, obtains top edge midpoint straight line Lup:yup=kupx+bup, to central point sequence
Row carry out least square fitting, obtain central point straight line Lmid:ymid=kmidx+bmid, a most young waiter in a wineshop or an inn is carried out to point sequence in lower edge
Multiply fitting, obtain lower edge midpoint straight line Ldown:ydown=kdownx+bdown, and this three straight lines are calculated according to straight slope
Horizontal tilt angle αup、αmid、αdown, calculation formula such as (3)：
αl=arctankl l∈{up,mid,down} (3)。
4. the method as described in claim 1, it is characterised in that step S3 includes：
S31, for top edge midpoint straight line, ask corresponding to average distance of the point sequence to top edge midpoint straight line in top edgeFor central point straight line, corresponding center point sequence is sought to the average distance of central point straight lineFor lower edge midpoint
Straight line, ask corresponding to average distance of the point sequence to top edge midpoint straight line in lower edgeCalculation formula such as (4)：
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mover>
<mi>d</mi>
<mo>&OverBar;</mo>
</mover>
<mi>l</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</msubsup>
<msubsup>
<mi>d</mi>
<mi>i</mi>
<mi>l</mi>
</msubsup>
</mrow>
<mi>M</mi>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>l</mi>
<mo>&Element;</mo>
<mo>{</mo>
<mi>u</mi>
<mi>p</mi>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
<mo>,</mo>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
<mo>}</mo>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
WhereinRepresent sequence PointlIn i-th point arrive straight line Ll:yl=klx+blDistance, calculation formula such as (5)：
<mrow>
<msubsup>
<mi>d</mi>
<mi>i</mi>
<mi>l</mi>
</msubsup>
<mo>=</mo>
<mfrac>
<mrow>
<mo>|</mo>
<msub>
<mi>k</mi>
<mi>l</mi>
</msub>
<mo>*</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<msub>
<mi>b</mi>
<mi>l</mi>
</msub>
<mo>|</mo>
</mrow>
<msqrt>
<mrow>
<msubsup>
<mi>k</mi>
<mi>l</mi>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msqrt>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
S32, the confidence level for defining top edge midpoint straight line, central point straight line and lower edge midpoint straight line are respectively λup、λmid、
λdown, calculation formula such as (6)
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>&lambda;</mi>
<mi>l</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msub>
<mover>
<mi>d</mi>
<mo>&OverBar;</mo>
</mover>
<mi>l</mi>
</msub>
</mfrac>
</mrow>
</mtd>
<mtd>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mo>{</mo>
<mi>u</mi>
<mi>p</mi>
<mo>,</mo>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
<mo>,</mo>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
<mo>}</mo>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
And the weight for defining top edge midpoint straight line, central point straight line and lower edge midpoint straight line is respectively wup、wmid、wdown, meter
Calculate formula such as (7)
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
</msub>
<mrow>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mrow>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
</msub>
<mrow>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>u</mi>
<mi>p</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>m</mi>
<mi>i</mi>
<mi>d</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mrow>
<mi>d</mi>
<mi>o</mi>
<mi>w</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
5. the method as described in claim 1, it is characterised in that the step S4 is specially：
The inclined angle alpha of top edge midpoint straight line, central point straight line and lower edge midpoint straight line is calculated in above-mentioned stepsup、αmid、
αdown, and respective weight wup、wmid、wdown, the optimum level angle of inclination for thus calculating car plate is：
α=wupαup+wmidαmid+wdownαdown (8)。
6. the method as described in claim 1, it is characterised in that the step S5 is specially：
With license plate image central pointFor pivot, the above-mentioned optimum level tilt angle alpha that is calculated is
The anglec of rotation, rotate license plate image, you can obtain final horizontal tilt correction result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710784790.0A CN107563330B (en) | 2017-09-04 | 2017-09-04 | Horizontal inclined license plate correction method in surveillance video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710784790.0A CN107563330B (en) | 2017-09-04 | 2017-09-04 | Horizontal inclined license plate correction method in surveillance video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107563330A true CN107563330A (en) | 2018-01-09 |
| CN107563330B CN107563330B (en) | 2020-10-16 |
Family
ID=60977953
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710784790.0A Active CN107563330B (en) | 2017-09-04 | 2017-09-04 | Horizontal inclined license plate correction method in surveillance video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107563330B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109584165A (en) * | 2018-11-30 | 2019-04-05 | 泰康保险集团股份有限公司 | A kind of antidote of digital picture, device, medium and electronic equipment |
| CN109871775A (en) * | 2019-01-22 | 2019-06-11 | 北京影谱科技股份有限公司 | A kind of the ice rink monitoring method and device of Behavior-based control detection |
| CN109919155A (en) * | 2019-03-13 | 2019-06-21 | 厦门商集网络科技有限责任公司 | A kind of the inclination angle antidote and terminal of text image |
| CN110309828A (en) * | 2019-06-27 | 2019-10-08 | 浙江工业大学 | A kind of inclination license plate antidote |
| CN111652230A (en) * | 2020-05-25 | 2020-09-11 | 浙江大华技术股份有限公司 | License plate recognition method, electronic device and storage medium |
| CN112598001A (en) * | 2021-03-08 | 2021-04-02 | 中航金城无人系统有限公司 | Automatic ship water gauge reading identification method based on multi-model fusion |
| CN112752158A (en) * | 2020-12-29 | 2021-05-04 | 北京达佳互联信息技术有限公司 | Video display method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104318233A (en) * | 2014-10-19 | 2015-01-28 | 温州大学 | Method for horizontal tilt correction of number plate image |
| CN104408451A (en) * | 2014-10-30 | 2015-03-11 | 安徽清新互联信息科技有限公司 | Least-square-method-based license plate correction method |
| US20160314563A1 (en) * | 2013-12-25 | 2016-10-27 | Grg Banking Equipment Co., Ltd. | Method for correcting fragmentary or deformed quadrangular image |
-
2017
- 2017-09-04 CN CN201710784790.0A patent/CN107563330B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160314563A1 (en) * | 2013-12-25 | 2016-10-27 | Grg Banking Equipment Co., Ltd. | Method for correcting fragmentary or deformed quadrangular image |
| CN104318233A (en) * | 2014-10-19 | 2015-01-28 | 温州大学 | Method for horizontal tilt correction of number plate image |
| CN104408451A (en) * | 2014-10-30 | 2015-03-11 | 安徽清新互联信息科技有限公司 | Least-square-method-based license plate correction method |
Non-Patent Citations (2)
| Title |
|---|
| 王宗跃等: "基于字符上下边缘的车牌校正方法", 《武汉理工大学学报》 * |
| 赵娟等: "基于Hough变换及可信度的车牌旋转角度检测", 《武汉轻工大学学报》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109584165A (en) * | 2018-11-30 | 2019-04-05 | 泰康保险集团股份有限公司 | A kind of antidote of digital picture, device, medium and electronic equipment |
| CN109871775A (en) * | 2019-01-22 | 2019-06-11 | 北京影谱科技股份有限公司 | A kind of the ice rink monitoring method and device of Behavior-based control detection |
| CN109919155A (en) * | 2019-03-13 | 2019-06-21 | 厦门商集网络科技有限责任公司 | A kind of the inclination angle antidote and terminal of text image |
| CN109919155B (en) * | 2019-03-13 | 2021-03-12 | 厦门商集网络科技有限责任公司 | Inclination angle correction method for text image and terminal |
| CN110309828A (en) * | 2019-06-27 | 2019-10-08 | 浙江工业大学 | A kind of inclination license plate antidote |
| CN111652230A (en) * | 2020-05-25 | 2020-09-11 | 浙江大华技术股份有限公司 | License plate recognition method, electronic device and storage medium |
| CN111652230B (en) * | 2020-05-25 | 2023-05-12 | 浙江大华技术股份有限公司 | License plate recognition method, electronic device and storage medium |
| CN112752158A (en) * | 2020-12-29 | 2021-05-04 | 北京达佳互联信息技术有限公司 | Video display method and device, electronic equipment and storage medium |
| US11490164B2 (en) | 2020-12-29 | 2022-11-01 | Beijing Dajia Internet Information Technology Co., Ltd | Method and electronic device for displaying video |
| CN112598001A (en) * | 2021-03-08 | 2021-04-02 | 中航金城无人系统有限公司 | Automatic ship water gauge reading identification method based on multi-model fusion |
| CN112598001B (en) * | 2021-03-08 | 2021-06-25 | 中航金城无人系统有限公司 | Automatic ship water gauge reading identification method based on multi-model fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107563330B (en) | 2020-10-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107563330A (en) | A kind of horizontal tilt car plate antidote in monitor video | |
| CN105005771B (en) | A kind of detection method of the lane line solid line based on light stream locus of points statistics | |
| CN103324930B (en) | A kind of registration number character dividing method based on grey level histogram binaryzation | |
| CN104318258B (en) | Time domain fuzzy and kalman filter-based lane detection method | |
| CN104129389B (en) | Method for effectively judging and recognizing vehicle travelling conditions and device thereof | |
| CN107025432B (en) | A kind of efficient lane detection tracking and system | |
| CN111563412B (en) | Rapid lane line detection method based on parameter space voting and Bessel fitting | |
| CN104102905B (en) | A kind of adaptive detection method of lane line | |
| CN107679436B (en) | Image correction method suitable for bending deformation two-dimensional code | |
| CN104077756B (en) | Direction filtering method based on lane line confidence | |
| CN103295420A (en) | Method for recognizing lane line | |
| CN109255350B (en) | New energy license plate detection method based on video monitoring | |
| CN104156731A (en) | License plate recognition system based on artificial neural network and method | |
| US9087263B2 (en) | Vision based pedestrian and cyclist detection method | |
| CN107066933A (en) | A kind of road sign recognition methods and system | |
| CN104809433B (en) | A kind of zebra line detecting method based on maximum stable region and stochastical sampling | |
| CN109344704B (en) | Vehicle lane change behavior detection method based on included angle between driving direction and lane line | |
| CN101629820A (en) | Road-edge detection | |
| CN109190483B (en) | Lane line detection method based on vision | |
| KR20130052614A (en) | Three-dimensional object detection device and three -dimensional object detection method | |
| CN104537342B (en) | A kind of express lane line detecting method of combination ridge border detection and Hough transformation | |
| CN108009522B (en) | Road detection method, device and terminal | |
| CN109034019B (en) | Yellow double-row license plate character segmentation method based on row segmentation lines | |
| CN112381870B (en) | Binocular vision-based ship identification and navigational speed measurement system and method | |
| CN104484680B (en) | A kind of pedestrian detection method of multi-model multi thresholds combination |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |